Logistic Regression - Predicted Probabilities (part 1)

I demonstrate how to calculate predicted probabilities and group membership for cases in a binary (a.k.a., binomial) logistic regression analysis. I do so through the menu driven approach, as well as an approach that involves the direct estimation of the predicted probabilities by using the unstandardized beta weights and intercept derived from the binomial logistic regression analysis.

[@TimNield888]

Isn't there's quite a significant mistake in this video? When you're in the Logistic Regression window, and you're inserting your variables, aren't you supposed to define the fact that your 'gender' variable is categorical. I didn't see you do this in the video? Any reason why?

[@Adrian-sv1mo]

If I understand correctly from 03:00 you explain a second method of the menu method explained before this moment

[@TheJebacina]

Hi, could you explain why we need to convert probabilities into odds? Why can´t we just use probabilities, they are between 0 and 1.

[@evoshindi]

Can you dhow/mention how we can predict probabilities of group membership in a multinomial logistic? My DV has 3 categories

[@TheGwaradaFam]

hello there can you assist me , i have run the logistic regression and my classification tables show no improvement in the full model when compared to the empty model i.e the classification rate is the same. The results are however statistically significant at all levels . does that mean that my model is useless and the results obtained are also useless? how can l modify my data to get an overall improvement in the classification rate?

[@mojazzz284]

what program are you using for this

[@sarahmckenzie2255]

SPSS